Machine Learning Applications to Sports Injury: A Review
Hanna Sigurdson
1,3 a
and Jonathan H. Chan
2,3 b
1
Engineering Science, University of Toronto, Toronto, Canada
2
School of Information Technology, King Mongkut’s University of Technology Thonburi, Bangkok, Thailand
3
Innovative Cognitive Computing (IC2), King Mongkut’s University of Technology Thonburi, Bangkok, Thailand
Keywords: Machine Learning, Sports Injury, Intrinsic Factors, Extrinsic Factors, Sports Injury Risk Factors, Artificial
Intelligence, Literature Review, Sports Psychology.
Abstract: As sports injuries increase in frequency in adolescents, and injuries in professional athletes create a
detrimental impact on the sports industry, research surrounding preventing sports injuries becomes more
prevalent. The mechanism for sports injury is well defined and includes intrinsic (age, psychology etc.) and
extrinsic risk factors (weather, training load etc.), and the inciting event. With the rise of machine learning
(ML), a variety of ML techniques have been applied to various sports injury aspects. The purpose of this work
is to assess the current applications of ML to sports injury and identify areas of growth by a systematic analysis
of applications to each injury element: intrinsic factors, extrinsic factors, and the inciting event. Current
underdeveloped areas are identified as: psychological effect, use of extrinsic factors, analysis of the inciting
event, and application of the action recognition ability of videos and wearable technology. Future technical
applications in these underdeveloped areas should be undergone to expand on and improve sports injury
prevention technology.
1 INTRODUCTION
1.1 Sports Injuries
Sports injuries occur across all levels, including
children, adolescents, recreational adults of various
age groups, and professional athletes. This has a
detrimental impact on multiple aspects of society.
Sports injuries in the pediatric and adolescent
population are increasing in frequency (Habelt et al.,
2011; Kerssemakers et al., 2009). Injures can cause
long term health impacts for young players (Maffulli
et al., 2010). Sports injuries can also have a
significant effect on the mental health of the
individual at all levels of play, affecting self-esteem
and overall performance (Arvinen-Barrow & Walker,
2013; Nippert & Smith, 2008).
In professional sports, an injury can have a
significant impact on the player’s career, team
performance and the sports industry (Brock &
Kleiber, 1994; Warnock, n.d.). An informative
example of the magnitude of the monetary cost is the
a
https://orcid.org/0000-0001-6572-2686
b
https://orcid.org/0000-0002-2384-0462
$12.4 million U.S. per team that sports injuries cost
in the top four professional soccer leagues in 2015
(Guest, n.d.).
Sports injures across the general population have
resulted in an enormous economic cost for direct
medical care, rehabilitation, lost wages, and national
productivity losses. In 1994 these factors led to an
estimated $224 billion U.S. cost in the United States
of America (U. Johnson, 2011a).
As a result of the abundance of evidence towards
the detrimental effects of sports injuries, significant
efforts have been placed into understanding the
etiology and identifying various risk factors for sports
injuries.
1.2 Machine Learning
The recent advancements of machine learning (ML)
have resulted in recommendations to implement ML
in sports medicine (Beal et al., 2019; Claudino et al.,
2019; Martin et al., 2021; Ruddy et al., 2019; Van
Eetvelde et al., 2021).
Sigurdson, H. and Chan, J.
Machine Learning Applications to Sports Injury: A Review.
DOI: 10.5220/0010717100003059
In Proceedings of the 9th Inter national Conference on Sport Sciences Research and Technology Support (icSPORTS 2021), pages 157-168
ISBN: 978-989-758-539-5; ISSN: 2184-3201
Copyright
c
 2021 by SCITEPRESS – Science and Technology Publications, Lda. All rights reser ved
157

ML is a branch of artificial intelligence that uses
data to build models that make informed predictions
(Martin et al., 2021). It has a large variety of
applications, one being image classification and
analysis, which can be widely applied in the medical
field. To train a model, large quantities of high-
quality data are required, which translates to requiring
an expert understanding of etiology when applied to
sports medicine.
The developing technologies in sports injury
prediction and other sports areas based on ML can be
used to prevent injuries, therefore addressing the
negative impacts of sports injuries discussed. This
justifies the importance of novel applications and
implementation of such technology.
1.3 Purpose
This work aims to assess the current applications of
ML to sports injury and identifies areas of growth and
opportunities for new applications based on the
current understanding of injury risk factors.
2 THE INJURY MECHANISM
The mechanism of injury has been the topic of
numerous works that propose models for sports injury
(Bittencourt et al., 2016; Meeuwisse, 1994;
Meeuwisse et al., 2007; van Mechelen et al., 1992). A
common approach to representing the injury
mechanism and sports injury aetiology is categorising
various injury risk factors into intrinsic and extrinsic,
and then including the occurrence of an inciting
event. One such example is Meeuwisse et al.’s
multifactorial model shown in Figure 1 (Meeuwisse,
1994; Meeuwisse et al., 2007). Of the risk factors for
injury that are distant from the outcome there are the
intrinsic factors, which are individual physical and
psychological characteristics, and the extrinsic
factors which are external factors (Lysens et al.,
1984). Generally, both intrinsic and extrinsic factors
can be classified into “modifiable” and “non-
modifiable” factors.
An example list of risk factors is as follows:
extrinsic non-modifiable factors are kind of sport,
level of sport, position, time of season, and weather;
extrinsic modifiable factors are equipment, playing
surface, playing time, rules, and time of day; intrinsic
non-modifiable factors are age, previous injury, and
sex; intrinsic modifiable factors are coordination,
fitness level, flexibility, participation in sport-specific
training, proprioception, psychological factors, and
strength (Habelt et al., 2011).
Finally, once the predisposed athlete has become
susceptible there is opportunity for the inciting event,
the mechanism of injury that is proximal to the
outcome.
Figure 1: A dynamic, recursive model of sports injury
developed by Meeuwisse et. al (Meeuwisse et al., 2007),
based on work done by Van Mechelen (van Mechelen et al.,
1992).
Bittencourt et al. agreed that due to the complex
nature of sports injury, future research requires
movement from isolated risk facture research to
injury pattern recognition caused by interactions in a
web of determinants (Bittencourt et al., 2016). This
provides further argument for application of ML to
sports injury prediction, because of its strength in
application to multifactorial predictions.
This work leverages the injury mechanism by
examining which factors at various stages of injury
have been the focus of ML applications.
3 LITERATURE REVIEW
This section analyzes sports injury risk factors by
reviewing the current understanding of sports injury
etiology at different sources of risk (intrinsic and
extrinsic) as well as different stages of injury (both
the risk factors and the inciting event). Furthermore,
it aims to understand where recently popularized ML
techniques have been applied and where they have yet
to be explored. A mind map summary of the results
from this review is shown in Figure 2.
3.1 Methodology
This work followed a systematic approach to review
the literature surrounding ML applications to sports
injury, while also addressing risk factors that have
been identified for sports injuries. The guidelines
followed for this review were:
icSPORTS 2021 - 9th International Conference on Sport Sciences Research and Technology Support
158

Figure 2: Mind map summary of machine learning applications to sports injuries.
1. The authors conducted searches on the databases
Google Scholar, PubMed, and Scopus.
2. Search terms included “machine learning sports
injury”, “deep learning sports injury”, “machine
learning sports”, “predict injury”, “sports injury
causes”, “sports injury aetiology”, “psychology
sports injury”, “tired” OR “fatigue” “sports
injury”, “extrinsic factors sports injury”, and
“intrinsic factors sports injury”.
3. Inclusion and exclusion criteria for sports injury
risk factors included that the works must relate to
sports, and not solely other injury types.
Furthermore, the work cannot focus only on injury
rehabilitation, as injury rehabilitation is outside of
the scope of this work.
4. Inclusion and exclusion criteria for ML
applications included that the work must relate to
sports; however, the work can apply to aspects of
sports other than injury. The works must address
some type of ML application.
5. Abstracts of relevant works were read and if the
work remained within the inclusion and exclusion
criteria, the work was read and included in the
review.
3.2 Intrinsic Factors
Each athlete has personal risk factors that increase
their likelihood of sports injury. These are referred to
as intrinsic factors and are the subject of the majority
of ML applications to sports injury prediction and
prevention. Generally, these factors are classified into
modifiable and non-modifiable factors, and the
following will provide an overview of these factors
and the applications of ML previously applied to
them.
3.2.1 Intrinsic Factors Summary
Various intrinsic factors have been identified through
numerous research studies. Some of these include:
nutrition (Close et al., 2019) female gender, age
greater than 24 years, a high body mass index, low
level of physical fitness at the commencement of a
training program, a past history of injury, leg length
discrepancy, neuromuscular control, core instability,
and many more (Chorba et al., 2010). This is only a
select list and does not include all factors.
Other common factors focused on in this work
are: psychological factors, stretching and warm up,
training load and fatigue.
3.2.2 Psychological Factors
Examining the psychological impact on sports injury
occurrence has been a well-established subject for
many sports medicine researchers and suggests a
meaningful path to injury prediction and prevention
(Heil, 1993).
Williams and Andersen proposed a model of
psychological antecedents of sports injury and
provided a basis for future models shown in Figure 3
(Andersen & Williams, 1988; U. Johnson, 2011a).
Machine Learning Applications to Sports Injury: A Review
159

Figure 3: An adapted version of Andersen and Williams’
1988 model and 1998 revised model, created by (Arvinen-
Barrow & Walker, 2013; Johnson, 2011a).
Arvinen-Barrow et al. found significant
occurrences of sports injuries impacted by factors
such as personality, anxiety, stress response, locus of
control, mental and emotional stress, major life
events, daily hassles and occurrence of sports injuries
(Arvinen-Barrow & Walker, 2013). Some evidence
for stressors from difficult relationships or
disagreements and inability to cope may predict
sports injuries, and discussion of these factors is
important before the start of the season (Gould &
Petlichkoff, n.d.).
Furthermore, attention to psychology related to
injury rehabilitation is very common, likely because
injury gives incentive to monitor an athlete's
psychological state (Brewer & Cornelius, 2003;
Concannon & Pringle, 2012). Fernandes et al.
suggested that there needs to be sufficient athlete
support after injury; however, in some situations,
psychological intervention may be detrimental and
requires educated coaches and rehabilitation
supporters (Fernandes et al., 2014). Psychological
impact on injury rehabilitation has also been
approached from the angle of the effect of “burnout”,
which can disturb the regulation of the central
nervous system, the autonomic nervous system, and
the neuroendocrine system. This effect may manifest
in terms of physical symptoms and/or behavioral
changes (Ahern & Lohr, 1997).
Among psychological analysis, gender is also
considered to be a factor that may influence sports
injury. Wiese-Bjornstal et al. observed that
perfectionistic, competitive, and compulsive
personality factors predicted greater frequency of
overuse injury in female collegiate athletes (Wiese-
Bjornstal et al., 2015). After extensive literature
review, they concluded that gender did play a role in
sports injury and is worth further investigation.
3.2.3 Stretching and Warm-up
Stretching and warm-up before and after exercise or
sports is widely recommended to prevent injury and
improve player performance (Thacker et al., n.d.).
Despite this shared standard for athletes at the
professional and recreational level, a literature review
conducted by Thacker et al. did not find sufficient
evidence to make conclusions that stretching and pre-
play warm ups could prevent injury, and suggested
further investigation in this area (Thacker et al., n.d.).
While they did however find some evidence for
extremes of inflexibility and hyper flexibility to
increase the risk of injury. Particular focus on
hamstring muscle injury, however, led Worrell and
Perrin to conclude that stretching, hamstring
flexibility, warm-up, strength and fatigue all can
affect hamstring injury and should be used to help
injury prevention (Worrell & Perrin, 1992).
Functional screening tools assessing intrinsic
factors of joint laxity, age, decreased range of motion
of hip abduction, height (knee ligament injuries), may
be good predictors of various injury types.
Specifically, ankle injuries could use predictors such
as greater strength of the plantar flexors, greater BMI,
and postural sway (Dallinga et al., 2012).
3.2.4 Fatigue
Player and muscle fatigue could produce higher
potential for sports injury (Fuller et al., 2016; Mclean
et al., 2007; Worrell & Perrin, 1992). This is also a
highly modifiable prevention mechanism due to the
short-term nature of fatigue. In rugby, Fuller et al.
found that there were higher incidences of injury
during the second half of the match, likely due to
player fatigue (Fuller et al., 2016). Further research
on NCAA athletes found that particularly in female
athletes fatigue “increased initial and peak knee
abduction and internal rotation motions and peak
knee internal rotation, adduction, and abduction
moments” (Mclean et al., 2007). Fatigue causing
reduced leg control could increase the risk of ACL
injury during landings. This supports that fatigue can
lead to higher injury risk.
3.3 Intrinsic Factors and ML
Of all the applications of ML to sports injuries, use of
intrinsic factors to predict future injuries was found to
be the most common application.
After review, a variety of ML algorithms have
been identified as being used for prediction of injury
based on intrinsic factors. Generally, the features
icSPORTS 2021 - 9th International Conference on Sport Sciences Research and Technology Support
160

selected can be separated into non-modifiable/longer-
term, modifiable/shorter-term, and a combination of
both.
3.3.1 Non-modifiable and Longer-term
Factors
Although works focusing solely on non-modifiable
intrinsic factors are rare, they have been able to
produce predictive models with Area Under the
Curve (AUC) values around 0.65. Studies focusing on
unmodifiable risk factors such as sex, knee joint
laxity, medial knee displacement, height and other
factors such as socioeconomic status were used as
predictors in adolescent basketball and floorball
players (Jauhiainen et al., 2021). A similar work
focused on adolescent athletes (Karuc et al., 2021) in
an attempt to respond to growing adolescent sports
injury. An application to dental injuries including
non-modifiable factors as well as history of dental
injury and socioeconomic class also produced
meaningful predictive ability (Farhadian et al., 2020).
In adult Australian footballers, non-modifiable
factors such as age, stature, mass, primary playing
position, and lower limb injury history achieved an 85
% accuracy using XGBoost, indicating that these
factors can produce relatively accurate injury
prediction (Ruddy et al., 2018).
3.3.2 Modifiable and Shorter-term Factors
Player fatigue and overuse of the body has been used
as an injury predictor, which reflects the literature
previously discussed that supports fatigue as an
intrinsic factor. Using a Subgroup Discovery
approach on 14 elite male volleyball players, risk
factors such as fatigue, overuse, sleep, muscle
soreness and training exertion produced positive
predictive results (Leeuw et al., 2021). A similar
application to professional soccer players using
predictors of high metabolic rate and sudden
decelerations using a Decision Tree predicted 58 %
of injuries (Rossi et al., 2017), indicating the
importance of fatigue measurement in injury
prediction.
The motion and mobility of legs or other parts of
the body has also been a popular measured value for
injury prediction. To predict ACL injuries in
basketball players, Taborri et al. used the features of
leg stability, leg mobility, and capability to absorb the
load after jump measured by inertial sensors and
optoelectronic bars and obtained a 96 % accuracy
using Support Vector Machine (Taborri et al., 2021).
Motion assessment (e.g. single leg hop distance, tuck
jump assessment etc.) in youth soccer players also
obtained reasonable injury predictability using a
Decision Tree model (Oliver et al., 2020).
Using wearable technology to predict injury has
seen some early implementation; for example, IMUs
and surface electromyography (sEMG) electrodes
were used to measure predictors such as obliquity of
the pelvis, fall of the contralateral pelvis, the
extension of the knee, dorsiflexion of the ankle in the
initial contact, and less activation of the gluteus
medium during the first phase of float in triathletes
(Martínez-Gramage et al., 2020). A Random Forest
model achieved an AUC of 0.8 for predicting injury
from this data, and Random Forest also succeeded
when predicting ACL reinjury using fine-grained
motion data, with an AUC value of 0.89 (Kim et al.,
2019). Future use of wearable technology during
movement and during the inciting event could be a
future step in predicting and understanding injury.
3.3.3 Combination of Modifiable and
Non-modifiable Factors
The logical progression for injury prediction is to
include both modifiable and non-modifiable intrinsic
factors. Physical attributes such as height and weight
used in conjunction with physical ability such as
strength, flexibility, speed, agility, and endurance in
youth soccer players achieved 85 % precision using
XGBoost (Rommers et al., 2020). The same
predictors applied to various NCAA athletes achieved
an AUC of 0.79 (Henriquez et al., 2020). From these
examples we can see the success of using both
modifiable factors and non-modifiable factors that
describe the physical condition of the athletes prior to
injury.
Another interesting example uses player and team
statistics with previous injury statistics in NHL
players to produce an AUC performance from
XGBoost of 0.956 (Luu et al., 2020). This leaves one
to question which features in those statistics indicate
injury, as they do not provide a direct correlation to
injury in etiology studies. This example points to the
value of attempting various data sources even if there
is not much evidence initially of the impact of the
feature.
Although it may be hypothesized that works
combining both modifiable and non-modifiable
factors as predictors would produce higher
accuracies, the accuracy of each model is highly
variable based on previous works. For example,
Lopez-Valenciano et al. used a variety of predictors
including position, current level of play, dominant
leg, age, body mass, stature, body mass index, sleep
quality and burnout as well as physical attributes like
Machine Learning Applications to Sports Injury: A Review
161

dynamic postural control, lower extremity joint
ranges of motion and more (López-Valenciano et al.,
2018). However, they only achieved an AUC value of
0.76, which is not an improvement over other models
using fewer features. Direct comparison between
models including specific features could provide
better insight into the preferred risk factors to
measure. However, due to the current literature that
has highly specified datasets and little broad
application across varying datasets, it is difficult to
compare models trained on different datasets. Less
fine-tuning to specific datasets in future works could
be useful to apply models in a broader context.
3.3.4 Discussion of Intrinsic Factors
Despite the overwhelming evidence of psychological
risk factors, there is a lack of ML applications to these
risk factors. This can be hypothesized to be a result of
the private nature and difficulty of testing
psychological aspects of an athlete.
Many of the works relating to intrinsic factors are
limited in the robustness of their datasets, sampling
only tens of athletes (Huang & Jiang, 2021; Leeuw et
al., 2021; Naglah et al., 2018; Taborri et al., 2021).
Some works that used larger datasets with hundreds
of athletes could produce higher accuracy values and
are more generalized, suggesting that this could be a
more useful approach when such data is available
(Luu et al., 2020; Rommers et al., 2020).
Furthermore, many seem to use arbitrary features,
choosing certain intrinsic factors but excluding others
even though there is sufficient evidence to support the
excluded features in injury prediction. More research
on the effect of using a broader scope of predictors is
an area that may improve the accuracy of current
injury predictive technologies.
3.4 Extrinsic Factors
Extrinsic or “external” factors are environmental
conditions that increase the risk of sports injury
(Andersen & Williams, 1988; Fuller et al., 2016).
An important extrinsic factor is the floor or field
conditions. An example of this was concluded in
(Olsen et al., 2003), where they found that artificial
floors resulted in higher injury risk than wooden
floors for female team handball players.
The level of play and type of sport can influence
sports injury. An informative example is that lower
performance levels in combat sports showed a higher
frequency of injuries, and in karate, newer
competitors are also significantly more prone to
injury (Hammami et al., 2018). Between combat
sports there was also a discrepancy in sports injury
risk based on the type of combat sport (e.g. highest
risk in mixed martial arts).
3.4.1 Training Load
A key component of athlete success is regulation of
their training schedule and load. Jones et al.
conducted a literature review summarizing the
understanding of impact of training load and fatigue
on injury and illness (Jones et al., 2017). One of their
conclusions verified that during periods of training
load intensification, accumulation of training load
and acute change in load can increase risk in injury
and modifying training load during these times could
aid in preventing injury or illness. Although some
conclusions were made, they also noted the difficulty
to make broad conclusions due to other intrinsic and
extrinsic factors such as fitness, body composition,
playing level, injury history and age (Jones et al.,
2017). Another example case study is prevalence of
patellar tendon injuries in elite male soccer players,
where multivariate logistic regression showed that
high total exposure hours and increased body mass
were risk factors (Hägglund et al., 2011).
3.5 Extrinsic Factors and ML
Despite the evidence of the effect of extrinsic factors
on sports injury risk, there are only a few applications
of ML to this aspect of sports injury to the extent of
the knowledge of this review.
Some evidence of application is focused on
training load, and the impact of playing or training
time on injury. One example used GPS tracking data
to describe the training load of 26 professional soccer
players, and then applied a Decision Tree model to
predict 80 % of the injuries with about 50 %
precision (Rossi et al., 2018). GPS training data in
soccer using an XGBoost model achieved 95 %
accuracy (Vallance et al., 2020), indicating the
advancements of this method in injury prediction.
Similar work using GPS data to assess exertion in
rugby players also produced promising results
(Thornton et al., 2017), which proves the ability of
application across multiple sports. Other works
included training hours or time in conjunction with
intrinsic factors to predict sports injuries (Naglah et
al., 2018; Song et al., 2021).
A sole example of using the playing surface
applied a proposed Artificial Neural Network to 21
soccer players, and included proposed
epidemiological predictors of interface, shoes,
contaminants, damage on the playing surface and the
icSPORTS 2021 - 9th International Conference on Sport Sciences Research and Technology Support
162

athlete's movement's nature (Huang & Jiang, 2021).
Future inclusion of these features could potentially
improve injury prediction.
The inclusion of weather as features during injury
prediction is not common but has shown promise:
when applied to soccer games (Landset et al., 2017)
and ski injuries (Radovanovic et al., 2019). Further
use of the weather in predictions for outdoor activities
is a promising avenue for future model features.
3.6 Inciting Event
The inciting event is considered the immediate cause
of injury and is often the extent of injury knowledge
for many people. Due to the short-term nature of the
occurrence, very little aspects of the inciting event
have seen ML application as compared to the intrinsic
factors. This review includes only one work relating
to the inciting event.
The early detection of injuries in Major League
Baseball Pitchers using video (Piergiovanni & Ryoo,
2019), is an informative example of the power of deep
learning to make predictions using computer vision.
Piergiovanni et al. used video framing baseball
pitchers and applied optical flow (a method used for
motion recognition). They then used a Convolutional
Neural Network (CNN) trained on the optical flow to
binarily classify pitches as an injury or not an injury.
Some injuries such as hamstring strains achieved high
predictive accuracies at around 0.98, while low
accuracy for finger blisters at 0.64. This is an
illustrative example of a ML application to further
understand and prevent the inciting event.
3.7 Rehabilitation
To narrow the scope of this review, works relating to
rehabilitation have been excluded. Some examples of
ML applied to sports injury rehabilitation include
using visual analysis or other means to assess
rehabilitation progress (Ba, 2020; Chen & Yuan,
2021; Edouard et al., 2020; Su, 2019).
3.8 Other Sports Applications
Apart from injuries, recently ML has made significant
progress in applications to other aspects of sports
(Advancing sports analytics through AI research,
n.d.). Some examples of this include tracking players,
predicting motion, analyzing shots, understanding
strategy and so on, see a mind map summary shown
in Figure 4. Although this area has been a topic of
discussion in previous reviews, they are limited in
their analytical conclusions (Rajšp & Fister, 2020) or
breadth of research (Richter et al., 2021), and if injury
is included they fail to address known injury risk
factors (Claudino et al., 2019).
One popular area of research is game outcome
prediction, which was excluded from this review due
to the large scope of the area and that it is less likely
to be translated into future sports injury application
than methods such as wearable technology and video
analysis. Using player and team statistics, such
prediction can achieve around 70 % accuracy, and is
currently being developed (Cao, 2012).
The following section provides a summary of the
current applications and provides insight on their
potential applications to sports injury.
3.8.1 Video Data Analysis
In recent years computer vision advancements have
led to success in analysis using video as input data. A
main challenge focus has been the identification of
actions in video and is an essential step towards injury
prediction through video. Recognizing tennis shots
(Mora & Knottenbelt, 2017; Skublewska-
Paszkowska et al., 2020), recognizing hockey and
soccer events (Vats et al., 2020), annotating tennis
events (de Castro, 2018), recognizing basketball
action (such as pass, shoot, catch, dribble) (Ji, 2020),
knee joint moment estimation (W. R. Johnson et al.,
2019), and automatic highlight generation (Lee et al.,
2020, p.) are applications identified in this review.
The use of video data to track multiple players is
an important application of ML to gain more
understanding of the influence of players and contact
in team sports on injuries. Evidence of this
technology is when Tian et al. used SportVU data
from the National Basketball Association teams to
achieve 69 % accuracy in defensive strategy
classification using a KNN model (Tian et al., 2020).
3.8.2 Wearable Technology Data Analysis
The use of inertial measurement units (IMU) to
collect data for action recognition is a common
application of ML in sports. This provides
opportunity for injury prevention, play enhancement
and coaching support (McGrath et al., 2020).
Similar to the developments in video, wearables
applied to a sports context are mainly focused on
activity recognition: general motion (Barshan &
Yüksek, 2014), basketball motions (Hu et al., 2020),
team handball throws and their speed (van den Tillaar
et al., 2021), kayakers’ strokes (Liu et al., 2021), and
tennis strokes (Sharma et al., 2017). Various models
were applied, for example: Artificial Neural Network,
Support Vector Machine (Gourgari et al., 2013),
Machine Learning Applications to Sports Injury: A Review
163

Figure 4: Mind map summary of other sports applications of ML.
Random Forest, and Gradient Boosting (van den
Tillaar et al., 2021).
Wearable data applied to sports injury has focused
on identification of previous injury, for example
anterior cruciate ligament injury in rugby players
(Tedesco et al., 2020). This leaves an opening for the
use of wearable data to predict injuries based on the
athlete motion, or to analyze the precursors to the
inciting event.
An important observation is the disparity in
predictive ability based on skill level, where data
collected on professional athletes achieves a higher
accuracy which could pose difficulty when applying
the same model to novice players. When applied to
experienced players, the classification accuracy of
basketball action (shooting, passing, dribbling, and
lay-up) was 0.85, while only 0.65 for inexperienced
players (Hu et al., 2020). This may impact the
application of any sort of action or injury recognition
to younger or novice players. Despite these players
being at risk of injury, the same video or wearable
technology data may not be able to achieve high
predictive accuracies because of their inconsistency
of play and is an area to consider in future works.
Finally, fatigue measured using wearable
technology is a useful development that could
improve the prediction of sports injury. Predicting
fatigue during various exercises such as squats (Jiang
et al., 2021), and predicting bicep muscle fatigue
(Elshafei & Shihab, 2021) are two examples that
predict fatigue of the athlete. As discussed in section
3.2.5, fatigue is a verified risk factor for injury, and
this could provide an effective method of measuring
fatigue to be used as a short-term predictor of sports
injury in future models.
4 CONCLUSIONS
As sports injuries become a growing concern in the
adolescent population and adversely affect
professional athletes, future application of ML to
improve injury prediction and prevention is
necessary. This review provides an in-depth
understanding of the current areas of deficiency in the
ML applications to sports injury. Main areas that are
currently the focus of ML application are prediction
based on modifiable and non-modifiable intrinsic
factors, with some work focusing on extrinsic factors
as well, and very little addressing the inciting event.
More applications to extrinsic factors, psychological
factors, and the inciting event would be meaningful
next steps. For example, leveraging advancements in
wearable technology and video data analysis in sports
could drastically reduce sports injury concern. Future
technologies that can more accurately predict and
prevent injuries can help achieve the ultimate goal of
this work and all works attempting to lessen the
impact of sports injuries.
ACKNOWLEGMENTS
Thank you to Anika Shenoy, Abnash Kaur Bassi,
Arjun Sharma, Connor Glossop, Serena Xinran Liu,
Shafinul Haque and Andy Wei Liu, all from
Engineering Science at the University of Toronto for
support and encouragements during the creation of
this work.
REFERENCES
Advancing sports analytics through AI research. (n.d.).
Deepmind. Retrieved 22 July 2021, from
https://deepmind.com/blog/article/advancing-sports-
analytics-through-ai
Ahern, D. K., & Lohr, B. A. (1997). Psychosocial Factors
In Sports Injury Rehabilitation. Clinics in Sports
Medicine, 16(4), 755–768. https://doi.org/10.1016/S02
78-5919(05)70052-1
Andersen, M. B., & Williams, J. M. (1988). A model of
stress and athletic injury: Prediction and prevention.
Journal of Sport & Exercise Psychology, 10(3), 294–
306.
Arvinen-Barrow, M., & Walker, N. (2013). The Psychology
of Sport Injury and Rehabilitation. Routledge.
icSPORTS 2021 - 9th International Conference on Sport Sciences Research and Technology Support
164

Ba, H. (2020). Medical Sports Rehabilitation Deep
Learning System of Sports Injury Based on MRI Image
Analysis. Journal of Medical Imaging and Health
Informatics, 10(5), 1091–1097. https://doi.org/10.1166/
jmihi.2020.2892
Barshan, B., & Yüksek, M. C. (2014). Recognizing Daily
and Sports Activities in Two Open Source Machine
Learning Environments Using Body-Worn Sensor
Units. The Computer Journal, 57(11), 1649–1667.
https://doi.org/10.1093/comjnl/bxt075
Beal, R., Norman, T. J., & Ramchurn, S. D. (2019).
Artificial intelligence for team sports: A survey. The
Knowledge Engineering Review, 34. https://doi.org/
10.1017/S0269888919000225
Bittencourt, N. F. N., Meeuwisse, W. H., Mendonça, L. D.,
Nettel-Aguirre, A., Ocarino, J. M., & Fonseca, S. T.
(2016). Complex systems approach for sports injuries:
Moving from risk factor identification to injury pattern
recognition—narrative review and new concept. British
Journal of Sports Medicine, 50(21), 1309–1314.
https://doi.org/10.1136/bjsports-2015-095850
Brewer, B. W., & Cornelius, A. E. (2003). Psychological
Factors in Sports Injury Rehabilitation. In
Rehabilitation of Sports Injuries: Scientific Basis (pp.
160–183). John Wiley & Sons, Ltd. https://doi.org/
10.1002/9780470757178.ch8
Brock, S. C., & Kleiber, D. A. (1994). Narrative in
Medicine: The Stories of Elite College Athletes’
Career-Ending Injuries. Qualitative Health Research,
4(4), 411–430. https://doi.org/10.1177/104973239400
400405
Cao, C. (2012). Sports Data Mining Technology Used in
Basketball Outcome Prediction. Dissertations.
https://arrow.tudublin.ie/scschcomdis/39
Chen, X., & Yuan, G. (2021). Sports Injury Rehabilitation
Intervention Algorithm Based on Visual Analysis
Technology. Mobile Information Systems, 2021,
e9993677. https://doi.org/10.1155/2021/9993677
Chorba, R. S., Chorba, D. J., Bouillon, L. E., Overmyer, C.
A., & Landis, J. A. (2010). Use of a Functional
Movement Screening Tool to Determine Injury Risk in
Female Collegiate Athletes. North American Journal of
Sports Physical Therapy : NAJSPT, 5(2), 47–54.
Claudino, J. G., Capanema, D. de O., de Souza, T. V.,
Serrão, J. C., Machado Pereira, A. C., & Nassis, G. P.
(2019). Current Approaches to the Use of Artificial
Intelligence for Injury Risk Assessment and
Performance Prediction in Team Sports: A Systematic
Review. Sports Medicine - Open, 5(1), 28.
https://doi.org/10.1186/s40798-019-0202-3
Close, G. L., Sale, C., Baar, K., & Bermon, S. (2019).
Nutrition for the Prevention and Treatment of Injuries
in Track and Field Athletes. International Journal of
Sport Nutrition and Exercise Metabolism,
29(2), 189–
197. https://doi.org/10.1123/ijsnem.2018-0290
Concannon, M., & Pringle, B. (2012). Psychology in sports
injury rehabilitation. British Journal of Nursing, 21(8),
484–490. https://doi.org/10.12968/bjon.2012.21.8.484
Dallinga, J. M., Benjaminse, A., & Lemmink, K. A. P. M.
(2012). Which Screening Tools Can Predict Injury to
the Lower Extremities in Team Sports? Sports
Medicine, 42(9), 791–815. https://doi.org/10.1007/
BF03262295
de Castro, F. B. F. (2018). Efficiently Scaling the
Annotation of Tennis Videos via Crowdsourcing.
Edouard, P., Verhagen, E., & Navarro, L. (2020). Machine
learning analyses can be of interest to estimate the risk
of injury in sports injury and rehabilitation. Annals of
Physical and Rehabilitation Medicine, 101431.
https://doi.org/10.1016/j.rehab.2020.07.012
Elshafei, M., & Shihab, E. (2021). Towards Detecting
Biceps Muscle Fatigue in Gym Activity Using
Wearables. Sensors, 21, 759. https://doi.org/10.3390/
s21030759
Farhadian, M., Torkaman, S., & Mojarad, F. (2020).
Random forest algorithm to identify factors associated
with sports-related dental injuries in 6 to 13-year-old
athlete children in Hamadan, Iran-2018 -a cross-
sectional study. BMC Sports Science, Medicine &
Rehabilitation, 12(1), 69. https://doi.org/10.1186/
s13102-020-00217-5
Fernandes, H. M., Machado Reis, V., Vilaça-Alves, J.,
Saavedra, F., Aidar, F. J., & Brustad, R. (2014). Social
support and sport injury recovery: An overview of
empirical findings and practical implications. Revista
de psicología del deporte, 23(2), 0445–0449.
Fuller, C. W., Taylor, A. E., & Raftery, M. (2016). Should
player fatigue be the focus of injury prevention
strategies for international rugby sevens tournaments?
British Journal of Sports Medicine, 50(11), 682–687.
https://doi.org/10.1136/bjsports-2016-096043
Gould, D., & Petlichkoff, L. M. (n.d.). Psychology Of
Sports Injuries. 6.
Gourgari, S., Goudelis, G., Karpouzis, K., & Kollias, S.
(2013). THETIS: Three Dimensional Tennis Shots a
Human Action Dataset. 2013 IEEE Conference on
Computer Vision and Pattern Recognition Workshops,
676–681. https://doi.org/10.1109/CVPRW.2013.102
Guest, S. A. P. (n.d.). SAP BrandVoice: The Crippling Cost
Of Sports Injuries. Forbes. Retrieved 22 July 2021, from
https://www.forbes.com/sites/sap/2015/08/11/the-crip
pling-cost-of-sports-injuries/
Habelt, S., Hasler, C. C., Steinbrück, K., & Majewski, M.
(2011). Sport injuries in adolescents. Orthopedic
Reviews, 3(2), e18. https://doi.org/10.4081/or.2011.e18
Hägglund, M., Waldén, M., Zwerver, J., & Ekstrand, J.
(2011). Epidemiology of patellar tendon injury in elite
male soccer players. British Journal of Sports
Medicine
, 45, 324. https://doi.org/10.1136/bjsm.20
11.084038.41
Hammami, N., Hattabi, S., Salhi, A., Rezgui, T., Oueslati,
M., & Bouassida, A. (2018). Combat sport injuries
profile: A review. Science & Sports, 33(2), 73–79.
https://doi.org/10.1016/j.scispo.2017.04.014
Heil, J. (1993). Psychology of sport injury (pp. xiv, 338).
Human Kinetics Publishers.
Henriquez, M., Sumner, J., Faherty, M., Sell, T., & Bent, B.
(2020). Machine Learning to Predict Lower Extremity
Musculoskeletal Injury Risk in Student Athletes.
Machine Learning Applications to Sports Injury: A Review
165

Frontiers in Sports and Active Living, 0.
https://doi.org/10.3389/fspor.2020.576655
Hu, X., Mo, S., & Qu, X. (2020). Basketball Activity
Classification Based on Upper Body Kinematics and
Dynamic Time Warping. International Journal of
Sports Medicine, 41(4), 255–263. https://doi.org/10.10
55/a-1065-2044
Huang, C., & Jiang, L. (2021). Data monitoring and sports
injury prediction model based on embedded system and
machine learning algorithm. Microprocessors and
Microsystems, 81, 103654. https://doi.org/10.1016/
j.micpro.2020.103654
Jauhiainen, S., Kauppi, J.-P., Leppänen, M., Pasanen, K.,
Parkkari, J., Vasankari, T., Kannus, P., & Äyrämö, S.
(2021). New Machine Learning Approach for Detection
of Injury Risk Factors in Young Team Sport Athletes.
International Journal of Sports Medicine, 42(02), 175–
182. https://doi.org/10.1055/a-1231-5304
Ji, R. (2020). Research on Basketball Shooting Action
Based on Image Feature Extraction and Machine
Learning. IEEE Access, 8, 138743–138751.
https://doi.org/10.1109/ACCESS.2020.3012456
Jiang, Y., Hernandez, V., Venture, G., Kulić, D., & K.
Chen, B. (2021). A Data-Driven Approach to Predict
Fatigue in Exercise Based on Motion Data from
Wearable Sensors or Force Plate. Sensors (Basel,
Switzerland), 21(4), 1499. https://doi.org/10.3390/s21
041499
Johnson, U. (2011a). Psychosocial antecedents of sport
injury, prevention, and intervention: An overview of
theoretical approaches and empirical findings.
International Journal of Sport and Exercise
Psychology, 5. https://doi.org/10.1080/1612197X.20
07.9671841
Johnson, W. R., Mian, A., Lloyd, D., & Alderson, J. (2019).
On-field player workload exposure and knee injury risk
monitoring via deep learning. Journal of Biomechanics.
https://doi.org/10.1016/j.jbiomech.2019.07.002
Jones, C. M., Griffiths, P. C., & Mellalieu, S. D. (2017).
Training Load and Fatigue Marker Associations with
Injury and Illness: A Systematic Review of
Longitudinal Studies. Sports Medicine, 47(5), 943–974.
https://doi.org/10.1007/s40279-016-0619-5
Karuc, J., Mišigoj-Durakovic, M., Šarlija, M., Markovic,
G., Hadžic, V., Trošt-Bobic, T., & Soric, M. (2021).
Can Injuries Be Predicted by Functional Movement
Screen in Adolescents? The Application of Machine
Learning. The Journal of Strength & Conditioning
Research, 35(4), 910–919. https://doi.org/10.1519/
JSC.0000000000003982
Kerssemakers, S. P., Fotiadou, A. N., de Jonge, M. C.,
Karantanas, A. H., & Maas, M. (2009). Sport injuries in
the paediatric and adolescent patient: A growing
problem. Pediatric Radiology, 39(5), 471–484.
https://doi.org/10.1007/s00247-009-1191-z
Kim, D., Mandalapu, V., Hart, J. M., Bodkin, S., Homdee,
N., Lach, J., & Gong, J. (2019). Poster Abstract:
Examining Cross-Validation Strategies for Predictive
Modeling of Anterior Cruciate Ligament Reinjury.
2019 IEEE/ACM International Conference on
Connected Health: Applications, Systems and
Engineering Technologies (CHASE), 27–28.
https://doi.org/10.1109/CHASE48038.2019.00019
Landset, S., Bergeron, M. F., & Khoshgoftaar, T. M.
(2017). Using Weather and Playing Surface to Predict
the Occurrence of Injury in Major League Soccer
Games: A Case Study. 2017 IEEE International
Conference on Information Reuse and Integration
(IRI), 366–371. https://doi.org/10.1109/IRI.2017.86
Lee, Y., Jung, H., Yang, C., & Lee, J. (2020). Highlight-
Video Generation System for Baseball Games. 2020
IEEE International Conference on Consumer
Electronics - Asia (ICCE-Asia), 1–4. https://doi.org/
10.1109/ICCE-Asia49877.2020.9277391
Leeuw, A.-W. de, Zwaard, S. van der, Baar, R. van, &
Knobbe, A. (2021). Personalized machine learning
approach to injury monitoring in elite volleyball
players. European Journal of Sport Science, 0(0), 1–10.
https://doi.org/10.1080/17461391.2021.1887369
Liu, L., Wang, H.-H., Qiu, S., Zhang, Y.-C., & Hao, Z.-D.
(2021). Paddle Stroke Analysis for Kayakers Using
Wearable Technologies. Sensors (Basel, Switzerland),
21(3), 914. https://doi.org/10.3390/s21030914
López-Valenciano, A., Ayala, F., Puerta, Jos. M., DE Ste
Croix, M. B. A., Vera-Garcia, F. J., Hernández-
Sánchez, S., Ruiz-Pérez, I., & Myer, G. D. (2018). A
Preventive Model for Muscle Injuries: A Novel
Approach based on Learning Algorithms. Medicine and
Science in Sports and Exercise, 50(5), 915–927.
https://doi.org/10.1249/MSS.0000000000001535
Luu, B. C., Wright, A. L., Haeberle, H. S., Karnuta, J. M.,
Schickendantz, M. S., Makhni, E. C., Nwachukwu, B.
U., Williams, R. J., & Ramkumar, P. N. (2020).
Machine Learning Outperforms Logistic Regression
Analysis to Predict Next-Season NHL Player Injury:
An Analysis of 2322 Players From 2007 to 2017.
Orthopaedic Journal of Sports Medicine, 8(9),
2325967120953404. https://doi.org/10.1177/23259671
20953404
Lysens, R., Steverlynck, A., van den Auweele, Y., Lefevre,
J., Renson, L., Claessens, A., & Ostyn, M. (1984). The
Predictability of Sports Injuries: Sports Medicine, 1(1),
6–10. https://doi.org/10.2165/00007256-198401010-
00002
Maffulli, N., Longo, U. G., Gougoulias, N., Loppini, M., &
Denaro, V. (2010). Long-term health outcomes of
youth sports injuries. British Journal of Sports
Medicine, 44(1), 21–25. https://doi.org/10.1136/
bjsm.2009.069526
Martin, R. K., Pareek, A., Krych, A. J., Kremers, H. M., &
Engebretsen, L. (2021). Machine learning in sports
medicine: Need for improvement. Journal of ISAKOS:
Joint Disorders & Orthopaedic Sports Medicine, 6(1),
1–2. https://doi.org/10.1136/jisakos-2020-000572
Martínez-Gramage, J., Albiach, J. P., Moltó, I. N., Amer-
Cuenca, J. J., Huesa Moreno, V., & Segura-Ortí, E.
(2020). A Random Forest Machine Learning
Framework to Reduce Running Injuries in Young
Triathletes. Sensors (Basel, Switzerland)
, 20(21),
E6388. https://doi.org/10.3390/s20216388
icSPORTS 2021 - 9th International Conference on Sport Sciences Research and Technology Support
166

McGrath, J., Neville, J., Stewart, T., & Cronin, J. (2020).
Upper body activity classification using an inertial
measurement unit in court and field-based sports: A
systematic review. Proceedings of the Institution of
Mechanical Engineers Part P Journal of Sports
Engineering and Technology, 235. https://doi.org/
10.1177/1754337120959754
Mclean, S. G., Felin, R. E., Suedekum, N., Calabrese, G.,
Passerallo, A., & Joy, S. (2007). Impact of Fatigue on
Gender-Based High-Risk Landing Strategies. Medicine
& Science in Sports & Exercise, 39(3), 502–514.
https://doi.org/10.1249/mss.0b013e3180d47f0
Meeuwisse, W. H. (1994). Assessing Causation in Sport
Injury: A Multifactorial Model. Clinical Journal of
Sport Medicine, 4(3), 166–170.
Meeuwisse, W. H., Tyreman, H., Hagel, B., & Emery, C.
(2007). A dynamic model of etiology in sport injury:
The recursive nature of risk and causation. Clinical
Journal of Sport Medicine: Official Journal of the
Canadian Academy of Sport Medicine, 17(3), 215–219.
https://doi.org/10.1097/JSM.0b013e3180592a48
Mora, S. V., & Knottenbelt, W. J. (2017). Deep Learning
for Domain-Specific Action Recognition in Tennis.
2017 IEEE Conference on Computer Vision and
Pattern Recognition Workshops (CVPRW), 170–178.
https://doi.org/10.1109/CVPRW.2017.27
Naglah, A., Khalifa, F., Mahmoud, A., Ghazal, M., Jones,
P., Murray, T., Elmaghraby, A. S., & El-baz, A. (2018).
Athlete-Customized Injury Prediction using Training
Load Statistical Records and Machine Learning. 2018
IEEE International Symposium on Signal Processing
and Information Technology (ISSPIT), 459–464.
https://doi.org/10.1109/ISSPIT.2018.8642739
Nippert, A. H., & Smith, A. M. (2008). Psychologic Stress
Related to Injury and Impact on Sport Performance.
Physical Medicine and Rehabilitation Clinics of North
America, 19(2), 399–418. https://doi.org/10.1016/
j.pmr.2007.12.003
Oliver, J. L., Ayala, F., De Ste Croix, M. B. A., Lloyd, R.
S., Myer, G. D., & Read, P. J. (2020). Using machine
learning to improve our understanding of injury risk
and prediction in elite male youth football players.
Journal of Science and Medicine in Sport, 23(11), 1044–
1048. https://doi.org/10.1016/j.jsams.2020.04.021
Olsen, O. E., Myklebust, G., Engebretsen, L., Holme, I., &
Bahr, R. (2003). Relationship between floor type and
risk of ACL injury in team handball. Scandinavian
Journal of Medicine & Science in Sports, 13(5), 299–
304. https://doi.org/10.1034/j.1600-0838.2003.00329.x
Piergiovanni, A. J., & Ryoo, M. S. (2019). Early Detection
of Injuries in MLB Pitchers from Video. 2019
IEEE/CVF Conference on Computer Vision and
Pattern Recognition Workshops (CVPRW), 2431–2438.
https://doi.org/10.1109/CVPRW.2019.00298
Radovanovic, S., Petrovic, A., Delibašić, B., & Suknović,
M. (2019). Ski Injury Predictions with Explanations
(pp. 148–160). https://doi.org/10.1007/978-3-030-
33110-8_13
Rajšp, A., & Fister, I. (2020). A Systematic Literature
Review of Intelligent Data Analysis Methods for Smart
Sport Training. Applied Sciences, 10(9), 3013.
https://doi.org/10.3390/app10093013
Richter, C., O’Reilly, M., & Delahunt, E. (2021). Machine
learning in sports science: Challenges and
opportunities. Sports Biomechanics, 0(0), 1–7.
https://doi.org/10.1080/14763141.2021.1910334
Rommers, N., Rössler, R., Verhagen, E., Vandecasteele, F.,
Verstockt, S., Vaeyens, R., Lenoir, M., D’hondt, E., &
Witvrouw, E. (2020). A Machine Learning Approach to
Assess Injury Risk in Elite Youth Football Players.
Medicine & Science in Sports & Exercise, 52(8), 1745–
1751. https://doi.org/10.1249/MSS.0000000000002305
Rossi, A., Pappalardo, L., Cintia, P., Fernández, J., Iaia, F.,
& Medina, D. (2017, September 1). Who is going to get
hurt? Predicting injuries in professional soccer.
Rossi, A., Pappalardo, L., Cintia, P., Iaia, F. M., Fernàndez,
J., & Medina, D. (2018). Effective injury forecasting in
soccer with GPS training data and machine learning.
PLOS ONE, 13(7), e0201264. https://doi.org/10.1371/
journal.pone.0201264
Ruddy, J. D., Cormack, S. J., Whiteley, R., Williams, M.
D., Timmins, R. G., & Opar, D. A. (2019). Modeling
the Risk of Team Sport Injuries: A Narrative Review of
Different Statistical Approaches. Frontiers in
Physiology, 10, 829. https://doi.org/10.3389/fphys.20
19.00829
Ruddy, J. D., Shield, A. J., Maniar, N., Williams, M. D.,
Duhig, S., Timmins, R. G., Hickey, J., Bourne, M. N.,
& Opar, D. A. (2018). Predictive Modeling of
Hamstring Strain Injuries in Elite Australian
Footballers. Medicine and Science in Sports and
Exercise, 50(5), 906–914. https://doi.org/10.1249/
MSS.0000000000001527
Sharma, M., Srivastava, R., Anand, A., Prakash, D., &
Kaligounder, L. (2017). Wearable motion sensor based
phasic analysis of tennis serve for performance
feedback. 2017 IEEE International Conference on
Acoustics, Speech and Signal Processing (ICASSP),
5945–5949. https://doi.org/10.1109/ICASSP.2017.79
53297
Skublewska-Paszkowska, M., Powroznik, P., & Lukasik, E.
(2020). Learning Three Dimensional Tennis Shots
Using Graph Convolutional Networks. Sensors, 20(21),
6094. https://doi.org/10.3390/s20216094
Song, H., xiu-ying Han, Montenegro-Marin, C. E., &
krishnamoorthy, S. (2021). Secure prediction and
assessment of sports injuries using deep learning based
convolutional neural network. Journal of Ambient
Intelligence and Humanized Computing, 12(3), 3399–
3410. https://doi.org/10.1007/s12652-020-02560-4
Su, Y. (2019). Implementation and Rehabilitation
Application of Sports Medical Deep Learning Model
Driven by Big Data. IEEE Access, 7, 156338–156348.
https://doi.org/10.1109/ACCESS.2019.2949643
Taborri, J., Molinaro, L., Santospagnuolo, A., Vetrano, M.,
Vulpiani, M. C., & Rossi, S. (2021). A Machine-
Learning Approach to Measure the Anterior Cruciate
Ligament Injury Risk in Female Basketball Players.
Sensors, 21(9), 3141. https://doi.org/10.3390/s21093
141
Machine Learning Applications to Sports Injury: A Review
167

Tedesco, S., Crowe, C., Ryan, A., Sica, M., Scheurer, S.,
Clifford, A. M., Brown, K. N., & O’Flynn, B. (2020).
Motion Sensors-Based Machine Learning Approach for
the Identification of Anterior Cruciate Ligament Gait
Patterns in On-the-Field Activities in Rugby Players.
Sensors, 20(11), 3029. https://doi.org/10.3390/s201
13029
Thacker, S. B., Gilchrist, J., Stroup, D. F., & Kimsey, C. D.
(n.d.). CLINICAL SCIENCES Clinical Investigations
The Impact of Stretching on Sports Injury Risk: A
Systematic Review of the Literature.
Thornton, H. R., Delaney, J. A., Duthie, G. M., &
Dascombe, B. J. (2017). Importance of Various
Training-Load Measures in Injury Incidence of
Professional Rugby League Athletes. International
Journal of Sports Physiology and Performance, 12(6),
819–824. https://doi.org/10.1123/ijspp.2016-0326
Tian, C., De Silva, V., Caine, M., & Swanson, S. (2020).
Use of Machine Learning to Automate the
Identification of Basketball Strategies Using Whole
Team Player Tracking Data. Applied Sciences, 10(1),
24. https://doi.org/10.3390/app10010024
Vallance, E., Sutton-Charani, N., Imoussaten, A.,
Montmain, J., & Perrey, S. (2020). Combining Internal-
and External-Training-Loads to Predict Non-Contact
Injuries in Soccer. Applied Sciences, 10, 5261.
https://doi.org/10.3390/app10155261
van den Tillaar, R., Bhandurge, S., & Stewart, T. (2021).
Can Machine Learning with IMUs Be Used to Detect
Different Throws and Estimate Ball Velocity in Team
Handball? Sensors, 21(7), 2288. https://doi.org/
10.3390/s21072288
Van Eetvelde, H., Mendonça, L. D., Ley, C., Seil, R., &
Tischer, T. (2021). Machine learning methods in sport
injury prediction and prevention: A systematic review.
Journal of Experimental Orthopaedics, 8(1), 27.
https://doi.org/10.1186/s40634-021-00346-x
van Mechelen, W., Hlobil, H., & Kemper, H. C. G. (1992).
Incidence, Severity, Aetiology and Prevention of Sports
Injuries. Sports Medicine, 14(2), 82–99.
Vats, K., Fani, M., Walters, P., Clausi, D. A., & Zelek, J.
(2020). Event detection in coarsely annotated sports
videos via parallel multi receptive field 1D
convolutions. 2020 IEEE/CVF Conference on
Computer Vision and Pattern Recognition Workshops
(CVPRW), 3856–3865. https://doi.org/10.1109/CVP
RW50498.2020.00449
Warnock, R. (n.d.). The Effect of Injuries on Player and
Team Performance: An Empirical Analysis of the
Production Function in the National Hockey League.
51.
Wiese-Bjornstal, D. M., Franklin, A. N., Dooley, T. N.,
Foster, M. A., & Winges, J. B. (2015). Observations
About Sports Injury Surveillance and Sports Medicine
Psychology among Female Athletes. Women in Sport
and Physical Activity Journal, 23(2), 64–73.
https://doi.org/10.1123/wspaj.2014-0042
Williams, J. M., & Andersen, M. B. (1998). Psychosocial
antecedents of sport injury: Review and critique of the
stress and injury model’. Journal of Applied Sport
Psychology, 10(1), 5–25. https://doi.org/10.1080/1041
3209808406375
Worrell, T. W., & Perrin, D. H. (1992). Hamstring Muscle
Injury: The Influence of Strength, Flexibility, Warm-
Up, and Fatigue. Journal of Orthopaedic & Sports
Physical Therapy, 16(1), 12–18. https://doi.org/10.25
19/jospt.1992.16.1.12
icSPORTS 2021 - 9th International Conference on Sport Sciences Research and Technology Support
168